Scour at Various Hydraulic Structures: Sluice gates, Submerged bridges, Low weirs

Transcription

1 5 th International Symposium on Hydraulic Structures Brisbane, Australia, June 2014 Hydraulic Structures and Society: Engineering Challenges and Extremes ISBN DOI: /uql Scour at Various Hydraulic Structures: Sluice gates, Submerged bridges, Low weirs B.W. Melville 1 1 Department of Civil and Environmental Engineering The University of Auckland Private Bag, Auckland NEW ZEALAND b.melville@auckland.ac.nz Abstract: Prediction of likely scour depths is an important aspect of the design of hydraulic structures. While detailed guidance is available for the scour design of some structures, e.g. bridge foundations, scant information is available for many other structures. Recent, as yet unpublished, research findings for a number of fluvial structure types are presented herein. The research findings span scour downstream from sluice gates, scour in the vicinity of low weirs, together with a particular aspect of scour at bridge foundations that has previously received minimal attention. In each case, new data are presented, together with analyses of the data and simple (preliminary) design relationships are determined from the data. A feature common to bridges and low weirs is the form of the dependence of scour depth on flow intensity for clear-water and live-bed scour conditions. The new data are derived from small-scale laboratory experiments for non-cohesive (alluvial) bed materials; the research findings are limited accordingly. Keywords: fluvial hydraulic structures, scour, predictive equations, weir, sluice gate, bridge 1. INTRODUCTION The major damage to fluvial hydraulic structures occurs during floods. The main cause of damage is scour of the river-bed. Scour is a natural phenomenon caused by the flow of water in rivers and streams. It occurs when the erosive capacity of the flow exceeds the ability of the sediment bed to resist it. Scour is most pronounced in alluvial materials, but can also occur in silty materials and in rock, especially deeply weathered rock. Despite much study, the principles of analysis of scouring are not well established for many structure types. There remains a parallel need for improved design guidance. The aim of this paper is to present some recent, as yet unpublished, research findings for scour at three different hydraulic structures, namely sluice gates, low submerged weirs and bridges where the bridge deck is partially or completely submerged. Preliminary recommendations for design, derived from the new research, are given. The research presented herein applies to scour in non-cohesive (alluvial) materials only. The data ranges used in the analyses are given, either in the text of the paper and/or in the plots of the relevant independent (dimensionless) parameters. Application of the findings beyond the range of the independent (dimensionless) parameters would be inappropriate. Figure 1 shows examples of each of these structure types. 2. SLUICE GATE SCOUR Scour downstream from a sluice gate is an example of scour due to submerged, turbulent horizontal jets. Such jets are two-dimensional (2D) when the width to thickness (or depth) ratio of the jet flow is large. The 2D horizontal jets can occur as submerged jets and as attached jets. The degree of submergence of the jet and the associated level of dissipation of the energy of the jet affect the scour capacity of the flow. A number of different equations have been suggested for estimation of the local scour depth due to 2D horizontal jets, including those by Valentin (1967), Altinbilek and Basmaci (1973), Ali and Lim (1986),

2 Breusers and Raudkivi (1991), Hoffmans (1998), Aderibigbe and Rajaratnam (1998), Lim and Yu (2002), Hopfinger et al. (2004), Dey and Sarkar (2006) and Melville and Lim (2013). Figure 1 Examples of hydraulic structures 2.1. Melville and Lim (2013) methodology Melville and Lim (2013) presented an analysis of all of the known data for local scour depth developed downstream from sluice gates, with or without protective aprons, leading to development of a new comprehensive prediction equation. The data sources were Tarapore (1956), Iwagaki et al. (1965), Valentin (1967), Basmaci (1971), Rajaratnam (1981), Rajaratnam and MacDougal (1983), Chatterjee et al. (1994), Lee (1995), Aderibigbe and Rajaratnam (1998), Lim and Yu (2002) and Dey and (2006). Subsequently, Liu (2013) presented data for the influence of a protective riprap layer on the local scour. The Melville and Lim (2013) method is described next. A definition diagram for local scour due to 2D horizontal jets is shown in Figure 2, which illustrates a typical situation of a local scour hole downstream from a sluice gate. The local scour is affected by the intensity of the jet flow, the size and grading of the bed sediment, the presence of an apron between the jet origin and the sediment bed, the tail water conditions and the effect of a protective riprap layer if present. Most existing scour equations use y j as the primary length scale for local scour depth; y j is used herein as well. Figure 2 Definition diagram for local scour downstream from sluice gates

3 The local scour depth y s can be expressed as,,,,, (1) The terms in (1) are: Fr j = V j /(gy j ) 0.5, which expresses the strength of the applied flow, in terms of jet velocity V j and jet thickness y j y t /y j, which expresses the effect of tail water depth y t D 50 /y j, which expresses the effect of the median size of the sediment D 50 L a /y j, which expresses the effect of apron length L a σ g, which expresses the effect of sediment gradation σ N r, which expresses the effects of riprap size D r, riprap thickness nd r and length of riprap layer L r, where n is the number of layers of riprap. The effect on scour depth of each of the parameters in (1) is now considered. For this purpose, (1) is written in the following form (2) where C is a coefficient and the K s are dimensionless multiplying factors expressing the influence on scour depth of sediment size (K D ), tail water depth (K yt ), sediment gradation (K ), apron length (K L ) and riprap characteristics (K r ). Plots of the data, showing the influence on scour depth of each of these parameters, independently of the other parameters, are given in Melville and Lim (2013) for all K-factors except K r. The ranges of data used in the analysis are given in Melville and Lim (2013). The plot for K r is given in Figure 4. The data ranges for the new data defining K r are D r /y j = 0 to 2, L r /y j = 0 to 20 and nd r /y j = 0 to 6. Figure 3 shows the influence of Fr j on y s. All data in the plot apply to L a = 0 and g 1.5. Amongst the data plotted, D 50 /y j varies from to 1.38, while y t /y j varies from 1 to 107. The straight line shown in Figure 3 has the equation 3 (3) Figure 3 The influence of jet Froude number on local scour depth Eqn. (3) is an envelope to most of the data, indicating that a value of C = 3 is an upper limit to scour depth and the local scour depths given by (3) can be considered to be maxima for particular values of the jet parameters (V j and y j ).

4 Envelopes curves were fitted to each K-factor data plot to conservatively express the effect of each parameter; the equations of these curves are presented in Table 1. Table 1 Equations for the various K-factors Factor Sediment size, D 50 /y j Apron length, L a /y j Tail water depth, y t /y j Equation (4) (5) (6) Sediment gradation, σ (7) Riprap protection, N r (8) Eqn. (4) is based on data for which influences of tail water depth, sediment gradation and platform length are absent, i.e. y t /y j >10, σ g <1.5 and L a = 0. In finer sediments (D 50 /y j < 0.6), scour depth is unaffected by sediment size. For larger values of D 50 /y j, scour depths reduce rapidly with increasing D 50 /y j. For alluvial sediments, the condition D 50 /y j < 0.6 is unlikely to be met in practice. The data from which (5) was derived are from Lim and Yu (2002) with Fr j varying from 2 to 4. The data are expressed in terms of K L, which is the ratio of y s /y j for a particular value of L a /y j to that for L a /y j = 0, i.e. in the absence of an apron. Increasing apron length reduces local scour depths, as expected, with apron length becoming relatively insignificant for large apron lengths. The influence of y t on y s, as described by (6), is based on data from Iwagaki et al.(1965) and Dey and Sarkar (2006). For y t /y j > 6, scour depth appears to be relatively independent of y t /y j. For low tail water levels, local scour depth increases with increasing tail water depth up to a certain tail water depth beyond which the scour is unaffected. However, the data are scattered and the trends shown are inconclusive. In particular, the trend for y t /y j < 6 is unproven, being based on very few data points. There are few data available to assess the influence of g. The data used are from Dey and Sarkar (2006), which have been used to determine (7) for the influence of K σ. These data have Fr j = 2.37, y t /y j = 9.7 and L a /y j = The influence of a riprap protective layer is expressed in terms of the riprap number, N r, given in Table 1. Eqn. (8), which defines K r in terms of N r, is based on new data by Liu (2013). The data are plotted in Figure 4 in which the solid line is (8). As riprap size, layer extent and thickness increase, the scour depth reduces as expected.

5 Figure 4 The influence of riprap layer characteristics on local scour depth 2.2. Local scour depth estimation The equation for prediction of local scour depth downstream from sluice gates is 3 (9) Figure 5 is a comparison of scour depth estimates using (9) with all the available data, with the exception of those for which a riprap layer was used and the Valentine (1967) data for which the apron length was not specified. Nearly all of the data are enveloped by (9). Figure 5 Comparison of measured scour depths with scour depths estimated from (9) For situations where a riprap protective layer is employed, the data by Liu (2013), together with those by Dey and Sarkar (2006) are plotted in Figure 6, in which scour depth predictions are again based on (9). The plot indicates that (9) leads to safe estimates of scour depth in most cases.

6 Figure 6 Comparison of measured scour depths with scour depths estimated from (9), for situations where a riprap layer is present Melville and Lim (2013) presented figures showing the result of applying most of the existing equations to the data set used in this paper. The values shown in Table 2 are the number of under-predictions and the number of over-predictions exceeding 300%. It is clear that none of the existing equations, other than Melville and Lim (2013) give consistently reliable predictions of local scour depth, when tested against the available laboratory data. This may be partly due to many of the existing equations not considering the effect of apron protection. The equations by Altinbilek and Basmaci (1973) and Aderibigbe and Rajaratnam (1998) have the advantage of few under-predictions, although the former leads to unrealistically high scour depth predictions in some cases. The Chatterjee et al. (1994) and Hopfinger et al.(2004) equations feature a significant number of under-predictions, as does the Hoffmans (1998) equation, which also has an inappropriate bias. The equations by Valentin (1967) and Breusers and Raudkivi (1991) show a consistent trend for most of the data, but have a significant number of relatively small under-predictions and a small number of significant over-predictions. The Lim and Yu (2002) and Dey and Sarkar (2006) equations also show a consistent trend for many of the data, but lead to many under-predictions. Table 2 Comparison of measured scour depths with scour depths predicted using existing equations Investigator Percentage of overpredictions greater than 300% Percentage of underpredictions Aderibigbe and Rajaratnam, Altinbilek and Basmaci, Breusers and Raudkivi, Chatterjee et al, Dey and Sarkar, Hoffmans, Hopfinger et al, Lim and Yu, Valentin, Melville and Lim,

7 2.3. Conclusions The findings of the recent research presented for local scour depth prediction downstream from sluice gates can be summarised as follows: The new data are derived from small-scale laboratory experiments for non-cohesive (alluvial) bed materials; the research findings are limited accordingly Local scour depth downstream from sluice gates can be estimated using factors for the effects of the jet characteristics, the tail water depth, the sediment size and gradation, the extent of a downstream protective apron, and the characteristics of a protective riprap layer The local scour depth can be reliably predicted using equation (9) The maximum possible local scour is given by equation (3) Equation (9) performs better than existing predictive equations when judged against the known laboratory data. A number of the existing predictive equations perform poorly when compared with the known data. 3. PRESSURE-FLOW SCOUR AT BRIDGES One aspect of bridge scour estimation technology that has received comparatively limited attention from researchers is the so-called pressure-flow scour, which can occur when the water surface elevation upstream from the bridge rises above the bridge low chord. As a consequence, the bridge is subject to a vertical contraction of the flow that increases the bridge opening velocity and potentially the sediment transport capacity. Orifice flow can occur through the bridge section if the flow does not overtop the bridge superstructure, or the bridge can be overtopped. Pressure-flow scour is caused by vertical constriction of the flow and should be distinguished from scour due to lateral flow contraction at the bridge crossing; the latter is termed contraction scour. Studies of pressure-flow scour at bridges include Abed (1991), Jones et al. (1993), Arneson (1997), Arneson and Abt (1998), Umbrell et al. (1998), Lyn (2008) and Guo et al. (2009). The latter measured scour depths for clear-water scour conditions only. An equation to predict pressure-flow scour is given in HEC-18 (Richardson and Davis, 2001), the method being based on the Arneson (1997) data for clear-water and live-bed transport scour conditions. Lyn (2007) pointed out unsatisfactory features of the HEC-18 equation, notably that the basic model of the original (Arneson, 1997) regression analysis suffers from a spurious correlation. Lyn (2007) proposed a new equation as follows ,0.5 (10) where y is approach flow depth, V is approach flow velocity in the bridge section before scour, V c is approach flow velocity at the threshold of sediment transport and y s is scour depth measured from the undisturbed bed level. A definition diagram is given in Figure 7, in which h b is bridge deck elevation and w is bridge overtopping flow depth. Amongst Arneson s data, the maximum recorded scour depth was equivalent to 0.5y. In this respect, it is noted that Arneson measured maximum scour depths, i.e. including the effects (trough depth) of passing bed-forms. This factor is considered in more detail below. Figure 7 Definition diagram for pressure-flow scour at bridges

8 3.1. New data The results from a recent study undertaken at the University of Auckland are now presented and then analysed together with some existing data. The new data were measured in a 440 mm wide sediment-recirculating flume under live-bed conditions for sediment transport. The data are given in Table 3. Table 3 New data for pressure-flow scour under live-bed transport conditions V (m/s) h b (mm) w (mm) y s (mm) y s /y V/V c h b /y The approach flow depth was y = 0.2 m. Sediment was a uniform 0.8 mm sand, with V c = 0.38 m/s. The bridge deck model spanned the flume; it was 208 mm wide and the height of the bridge superstructure (y-h b -w) was 55 mm. Scour depths were measured using an array of 24 SeaTek transducers positioned in-line beneath the bridge deck and upstream and downstream from the bridge model. The latter data were used to investigate bed-form characteristics upstream and downstream from the bridge, these data not being presented herein. The transducers were moved laterally in 40 mm increments, allowing the bed levels throughout the scour zone to be recorded. Scour depths for each transducer were taken as temporal averages to eliminate the effect of bed-forms. Experimental durations were selected to ensure that the sediment bed was in a state of dynamic equilibrium. The values of y s given in Table 3 are the maximum temporal average values for the data measured over the area beneath the bridge deck; these values are used for analysis. Typically, the deepest scour occurred about 150 mm downstream from the upstream edge of the bridge deck and the temporal average scour depths showed minimal variation across the flume Data analysis The pressure-flow scour depth, y s, depends on the applied flow, the bed sediment characteristics and the bridge geometry. For steady, uniform conditions, the applied flow can be expressed in terms of the average approach flow velocity (V) and depth (y); the effects of the bed sediment on the sediment erosion process can be represented by the critical flow velocity for sediment entrainment (V c ); while the effect of the bridge geometry on pressure-flow scour can be expressed in terms of bridge elevation (h b ). The depth of submergence (w) is not included in the analysis, because y was a constant for the

9 dataset, i.e. an increase in h b corresponds to a decrease in w, and vice-versa. Also, the pressureflow scour is expected to be significantly more dependent on the contracted flow beneath the bridge deck, than on the overtopping flow. Thus, y s can be written in the following simplified form or, in dimensionless form,,, (11), (12) where V/V c is the flow intensity, representing the capacity of the flow to scour the sediment bed, and h b /y represents the degree of vertical flow contraction at the bridge section. Eqn. (12) includes the variables of a simple vertical contraction scour equation, i.e. (y s + h b )V c = f(yv). Figure 8 shows the influence of flow intensity on pressure-flow scour depth at submerged bridges. Two sets of data are plotted, the new live-bed data data and data from Umbrell et al. (1998) for clearwater scour. The former are plotted with hollow symbols, while the latter are plotted with solid symbols. In the figure, the same symbol (hollow or solid) signifies the same, or very similar, values of h b /y as shown in the legend. The trends depicted resemble those for local scour at bridge foundations. Pressure-flow scour is initiated at about V = 0.5V c for lower deck elevations (h b /y = 0.5), this value increasing with increasing h b /y. At h b = y, it is expected that pressure-flow scour would not occur. For a particular h b /y, pressure-flow scour, under clear-water conditions, increases rapidly with increasing flow velocity to a peak value at V = V c. Under live-bed conditions, the scour depth decreases from the threshold velocity peak as sediment is transported into the bridge section from upstream. There appears to be a second peak in the scour profile at about V = 2V c, this peak being analogous to the live-bed peak for local scour at bridges. The relationship between pressure-scour depth and its independent parameters, flow intensity (V/V c ) and bridge deck elevation (h b /y), is a family of curves for different values of h b /y, with scour depth decreasing with the latter. Figure 8 Effect of flow intensity on pressure-flow scour at bridges Figure 9 shows the influence of bridge deck elevation on pressure-flow scour. The data plotted are the new live-bed scour data, together with two sets of data from Umbrell et al. (1998) for clear-water scour.

10 Figure 9 Effect of bridge deck elevation on pressure-flow scour at bridges The expected limit for pressure-flow scour at h b = y, as discussed above, is shown in the figure. The dashed line for h b /y < 0.4 indicates the expected lower limit for decreasing bridge deck elevation, i.e. an increase in scour depth with decreasing h b /y towards a peak value at relatively low bridge deck elevations, with a rapid decrease in scour depth thereafter, as h b tends to zero. The dashed lines for h b /y > 0.4 are indicative of the effect of varying V/V c on scour depth Predictive equation The new data are plotted with those of Arneson (1997) and Umbrell et al. (1998) in Figure 10. Figure 10 Available data for pressure-flow scour at bridges The figure includes 152 scour measurements, representing all of the new data, all of Umbrell et al. (1998) data and all of the Arneson (1997) data except a few data showing negative scour depths and

11 Arneson s live-bed data which appear to include bed-form heights (i.e. dune trough levels were apparently recorded). The equation of the solid (envelope) line is (13a) (13b) This equation can be used to give maximum likely pressure-flow scour depths for design purposes. Local scour, general scour and bed-form effects need to be added to pressure-flow scour to get total scour. As additional research findings become available, a factor can be added to the equation for pressure-flow scour depth for the effect of h b /y, and possibly other effects also. It is noted that (10) and (13) are similar Conclusions The findings of the recent research presented for pressure-flow scour depth prediction at partially or fully submerged bridges can be summarised as follows: The new data are derived from small-scale laboratory experiments for non-cohesive (alluvial) bed materials; the research findings are limited accordingly Excluding the effects of dune trough depths, pressure-flow scour depths measured in smallscale laboratory experiments are less than 45% of the approach flow depth Pressure-flow scour depth at partially or fully submerged bridges can be estimated using (13) The form of the dependence of scour depth on flow intensity, for clear-water and live-bed scour conditions, is similar to the well-established function for local scour at bridge foundations Additional data are needed to establish the influence of other independent parameters on pressure-flow scour depth. 4. SCOUR AT SUBMERGED WEIRS Weirs or sills are low head hydraulic structures across channels for the purposes of limiting excessive bed degradation, bed stabilization, raising upstream water level and reducing flow velocity (figure 11). The subject of scour downstream of bed sills has been extensively studied and many scour prediction equations have been published including Bormann and Julien (1991), Gaudio et al. (2000), Lenzi et al. (2003), D'Agostino and Ferro (2004) and Marion et al. (2006). Other recent studies include Ben Meftah and Mossa (2006) and Pagliara and Kurdistani (2013). However, most existing equations were developed for unsubmerged weirs or partially submerged sills, for which downstream scouring is a result of free overfall plunging (or partially submerged impinging) jets. Limited research has been reported for scouring at submerged weirs under live-bed conditions. New experimental results from a study on the scouring process at submerged weirs under live bed scour conditions are presented. Figure 11 Definition diagram for local scour at low submerged weirs.

12 For a fully submerged weir, Wu and Rajaratnam (1996) classified flow regimes over the weir as (1) surface jet; (2) surface wave; (3) breaking wave (or surface jump); and (4) impinging jet. Regimes (1) and (2), which might be collectively named the surface flow regime, feature the flow remaining as a jet at the surface in the downstream channel, with its thickness increasing downstream because of turbulent mixing (Wu and Rajaratnam, 1998). Regime (3) represents the transition from a surface jet regime to a plunging jet regime. For the impinging jet regime, the flow over the weir plunges into the tail water, diffuses as a plane submerged jet and eventually impacts the bed of the downstream channel. Previous research (Wu and Rajaratnam, 1996, 1998 and Ohtsu et al, 1997) indicates that the flow regimes are dependent on the head difference across the weir. The four cases are illustrated in Figure 12 (in the order vertically from surface jet to impinging jet), which includes photographs taken during this study Experiments Results from an on-going study of live-bed scour at low, submerged weirs are presented. The experiments were conducted in a tilting flume, 12m long, 0.44m wide and 0.58m deep. The flume has two pumps for recirculating both sediment and water. The sediment used in the experiments was a uniform, coarse silica sand, with a median diameter, d50 = 0.85 mm, while the tail water depth (yt) was controlled to be 150 mm. The weirs used in the experiments spanned the flume width and were manufactured from 10 mm thick rectangular plastic plates. The scour development around the weir and the bed elevation changes in the approach flow were measured as a function of time throughout the experiment using a Seatek Multiple Transducer Array (MTA). The instrument is an ultrasonic ranging system, comprising 64 transducers, which can detect the distance from the sensors to reflective objects. The measuring accuracy of the system is approximately ± 1 mm. Six clear-water tests and 42 live-bed tests were undertaken. For this study, the instrument operated at frequency between 1 Hz and 3 Hz, depending on the scour rate. Scour depths were measured immediately upstream from, and downstream from, the weir. (a) (b) (c) (d) Figure 12 Flow regimes for flow at submerged weirs; a) surface jet, b) surface wave, c) transition regime (breaking wave), and d) impinging jet

13 The upstream scour hole was deepest at the weir face. Three transducers, distributed across the flume immediately upstream from the weir, were used to determine the maximum scour. For the downstream scour measurements, 35 transducers were distributed across the scour measurement area, being an area spanning 480 mm in the flow direction and the full width of the flume. For each experiment, the upstream edge of the scour measurement area was moved longitudinally to a point between 300 mm and 500 mm downstream from the weir, to ensure it covered the deepest scour zone for that experiment. Figure 13 shows a typical set of results for the upstream (upper figure) and downstream (lower figure) scour holes, for one of the live-bed experiments. The datum for level measurements was the initial flat sediment bed. For the experiment shown in Figure 13, tail water depth (y t ) was 150 mm, approach flow velocity (V) was 0.77 m/s, weir height (z) was 30 mm and the experiment lasted 9.2 hours. The plot shows a temporal record of the maximum instantaneous scour depth within the measurement zone, i.e. the 35 records for the downstream scour hole, and the 3 records for the upstream scour hole, were interrogated to find the deepest scour at any time and these values were then plotted for all recording times throughout the experiment. A slightly different analysis was done for each experiment to determine the upstream scour depths, using the 3 transducers positioned immediately upstream from the weir. In this case, the maximum average scour depth was determined by averaging the peaks of the fluctuating variation in scour depth due to the passage of successive bedforms (as shown by the solid circles for the first 2 hours of the run). For the downstream scour hole results shown in Figure 13, the instantaneous maximum scour depth (shown by the solid circle) was 107 mm below the level of the initial flat bed, while the average maximum scour depth (y sd ) was 60 mm, as indicated by the dashed line. The latter was determined by averaging the scour depth record shown in Figure 13 over the 9.2 hour record, as was done for all experimental data for both the upstream and downstream scour holes. Figure 13 Typical experimental result for the temporal development of instantaneous maximum scour upstream (upper figure) and downstream (lower figure) from the weir

14 4.2. Results Figure 14 shows the average maximum scour depth, y su, upstream from the weir, as a function of flow intensity, V/V c. For this and other results, scour depths are normalised using tail water depth (a constant for the experiments). For live bed scour conditions, the observed scour and fill process upstream of the weir is induced by the approaching periodic bedforms, which were measured using transducers positioned at several positions upstream from the weir, coupled with the strong helical flow structures induced by the presence of the weir. For each weir height, the normalized scour depth increases with flow intensity to a maximum value at about V/V c = 2.2, and then decreases at higher flow intensities. Higher weirs induce shallower scour holes upstream from the weir, as shown in Figure 14. This is due to aggradation occurring upstream from the weir, with deeper aggradation for higher weirs at particular flow intensity conditions. Figure 14 Average maximum scour depth upstream of the weir, as a function of flow intensity A plot of bedform steepness, η/λ, against V/V c is shown in Figure 15. The bedform steepness for the approach flow is independent of weir height, as expected. Figure 15 Steepness of bedforms in the approach flow, as a function of flow intensity

15 Similar trends are depicted in Figures 14 and 15, suggesting that the migration of bedforms past the weir strongly influences the scour and fill process upstream from the weir. This is expected, because steeper dunes (and larger amplitude dunes) induce deeper scour with the arrival of the dune trough at the weir and correspondingly shallower scour with passage of the subsequent dune crest. The average maximum scour depth downstream from the weir is shown in Figure 16 as a function of flow intensity. Scour depth increases rapidly under clear water conditions, reaching a maximum value at the threshold condition of general sediment transport (V/V c = 1). Under live bed conditions, the scour depth decreases with increasing flow intensity as sediment is transported into the scour region from upstream. The minimum scour depth appears to occur near the flow regime transition from the surface flow regime to the impinging jet regime, as discussed above. Beyond the transition, scour depths rise steeply again. Figure 16 Average maximum scour depth downstream of the weir, as a function of flow intensity A similar trend to that shown in Figure 16 also occurs for local scour at bridge foundations, and for pressure-flow scour (as discussed in Section 3). At bridge foundations, scour depth initially reduces with increase in approach flow velocity, reaches a minimum value, and then increases again toward a second maximum. The second maximum occurs at about the transitional flat-bed stage of sediment transport on the channel bed and is termed the live bed peak (Melville, 1992, 1984). Although the trends are remarkable similar, whether the scour depth downstream of the weir reaches a live bed peak at higher flow intensities remains unproven due to insufficient experimental data and different scour mechanisms. Nevertheless, it is most likely that the scour depth will attain a second peak at high flow intensities. Additional research is being conducted to extend the range of data available under live bed conditions Predictive equations Following Section 3.2, the scour depth at a weir, y s, depends on the applied flow, the bed sediment characteristics and the weir geometry. For steady, uniform conditions, the applied flow can be expressed in terms of the average approach flow velocity (V) and tail water depth (y t ); the effects of the bed sediment on the sediment erosion process can be represented by the critical flow velocity for sediment entrainment (V c ); while the effect of the weir geometry on scour can be expressed in terms of weir elevation above undisturbed bed level (z). Thus, y s can be written in the following simplified form or, in dimensionless form,,, (14), (15)

16 where the scour depth, y s, is expressed as y su or y sd, for the upstream and downstream scour holes, respectively. For the study presented, the data ranges are 1.0 V/V c 3.1 and 0.2 z/y t Upstream scour depth The lines shown in Figure 14 are a good fit to the data for the average maximum scour depth upstream from the weir under live bed scour conditions. The lines have the following equation (16) Because the scour and fill process upstream from the weir is dependent on the approaching bedforms, y su can also be expressed in terms of bedform steepness, i.e. The following equation is a good fit to the data, (17) (18) Downstream scour depth The dashed lines shown in Figure 15 fit the data well. However, it would be inappropriate to use them for design due to the incompleteness of the data set on which they are based. The solid line shown in the figure envelopes all data for V/V c < 1. The equation of the solid line is (19a) (19b) This equation is conservative for 1 < V/V c < 2. At higher flow intensities, the data indicate that scour depths are likely to exceed those given by (19b). As noted above, the current project at The University of Auckland is on-going and other influences are being investigated, including the effects of sediment size, weir shape and tail water depth Conclusions The findings of the recent research presented for scour depth prediction upstream and downstream from low submerged weirs can be summarised as follows: The new data are derived from small-scale laboratory experiments for non-cohesive (alluvial) bed materials; the research findings are limited accordingly Maximum average scour depth upstream from low submerged weirs can be estimated using equations (16) or (18) Preliminary estimates of maximum average scour depth downstream from low submerged weirs can be estimated using (19) The form of the dependence of scour depth on flow intensity, for clear-water and live-bed scour conditions, is similar to the well-established function for local scour at bridge foundations Additional data are needed to augment the existing data and to establish the influence of other independent parameters on scour depth at weirs.

17 5. ACKNOWLEDGMENTS This work was undertaken by a number of current postgraduate students at the University of Auckland. In this respect, I wish to specifically acknowledge the excellent research undertaken by Dawei Guan and Cheng Liu. 6. REFERENCES Abed, L.M. (1991), Local scour around bridge piers in pressure flow, PhD thesis, Colorado State University, Fort Collins, Colorado, US. Aderibigbe, O. and Rajaratnam, N. (1998), Effect of sediment gradation on erosion by plane turbulent wall jets, Journal of Hydraulic Engineering 124 (10), Ali, K. and Lim, S.Y. (1986), Local scour caused by submerged wall jets, Proc.ICE, Part 2, 81, pp Altinbilek, H.D. and Basmaci, Y. (1973), Localized scour at the downstream of outlet structures, in: Proceedings of the 11th Congress on Large Dams. Madrid, pp Arneson, L.A. (1997), The effects of pressure flow on local scour in bridge openings, PhD. Thesis, Colorado State Univ., Fort Collins, Colorado, USA. Arneson, L.A. and Abt, S.R. (1998), Vertical contraction scour at bridges with water flowing under pressure conditions. Transportation Research Record, 1647, Basmaci, Y. (1971), Localised Scour Under Vertical Gates by Submerged Jets, M.Sc. Thesis, Middle East Tech. Univ. Turkey. Ben Meftah, M. and Mossa, M. (2006), Scour holes downstream of bed sills in low-gradient channels, Journal of Hydraulic Research, 44(4), Bormann, N. E. and Julien, P. Y. (1991), Scour Downstream of Grade-Control Structures, Journal of Hydraulic Engineering, 117(5), Breusers, H.and Raudkivi, A.J. (1991), Scouring, IAHR Hydraulic Structures Design Manual, AA Balkema. Chatterjee, S., Ghosh, S.and Chatterjee, M. (1994), Local scour due to submerged horizontal jet, Journal of Hydraulic Engineering, 120 (8), D'Agostino, V. and Ferro, V. (2004), Scour on Alluvial Bed Downstream of Grade-Control Structures, Journal of Hydraulic Engineering, 130(1), Dey, S.and Sarkar, A. (2006), Scour downstream of an apron due to submerged horizontal jets, Journal of Hydraulic Engineering 132, Gaudio, R. et al. (2000), Morphological effects of bed sills in degrading rivers, Journal of Hydraulic Research, 38(2), Guo, J., Kerenyi, K. and Pagan-Ortiz, J.E. (2009), Bridge pressure flow scour for clear water conditions, Technical Report, FHWA-HRT , United States Department of Transportation, Washington, D.C., 57 pp. Hoffmans, G.J.C.M. (1998), Jet scour in equilibrium phase, Journal of Hydraulic Engineering 124, Hopfinger, E.J., Kurniawan, A., Graf, W.H. and Lemmin, U. (2004), Sediment erosion by Görtler vortices: the scour-hole problem, Journal of Fluid Mechanics 520, Iwagaki, Y., Tsuchiya, Y., Imamura, M. (1965), Studies of the local scour from flows downstream of an outlet (1), Disaster Prevention Research Institute, Kyoto University Annuals No. 8, Jones, J.S., Bertoldi, D.A. and Umbrell, E.R. (1993), Preliminary studies of pressure-flow scour, ASCE Conf. on Hydraulic Engineering, ASCE, Reston, Virginia, US. Lee, W.W.L. (1995), Erosion downstream of a sluice gate, Final Year Project, Nanyang Technological University, Singapore. Lenzi, M. A. et al. (2003), Interference processes on scouring at bed sills, Earth Surface Processes and Landforms, 28(1), Lim, S. Y. and Yu, G. (2002), Scouring downstream of sluice gate, Proc. First International Conference on Scour of Foundations, Texas USA, Vol 1,

18 Liu, C. (2013), The effect of riprap protection on the scour process downstream of a sluice gate, M.E. Thesis, The University of Auckland, Auckland, New Zealand. Lyn, D.A. (2008), Pressure-flow scour: A re-examination of the HEC-18 equation, Journal of Hydraulic Engineering, 134 (7), Marion, A. et al. (2006), Sediment supply and local scouring at bed sills in high-gradient streams, Water Resources Research, 42(6), W Melville, B. W. (1984), Live bed Scour at Bridge Piers, Journal of Hydraulic Engineering, 110(9), Melville, B. W. (1992), Local Scour at Bridge Abutments, Journal of Hydraulic Engineering, 118(4), Melville, B.W. and Lim, S.Y. (2013), Scour caused by 2D horizontal jets, Journal of Hydraulic Engineering, 140(2) , doi: /(ASCE)HY Ohtsu, I. et al. (1997), Discussion of Submerged Flow Regimes of Rectangular Sharp-Crested Weirs by S. Wu and N. Rajaratnam, Journal of Hydraulic Engineering, 123(10), Pagliara, S., and Kurdistani, S. M. (2013), Scour downstream of cross-vane structures, Journal of Hydro-Environment Research (0). Rajaratnam, N. (1981), Erosion by plane turbulent jets, Journal of Hydraulic Research 19, Rajaratnam, N. and Macdougall, R.K. (1983), Erosion by plane wall jets with minimum tailwater, Journal of Hydraulic Engineering 109, Richardson, E.V. and Davis, S.R. (2001), Evaluating scour at bridges: Fourth edition, HEC-18, FHWA- NHI , United States Department of Transportation, Washington, D.C. Tarapore, Z.S. (1956), Scour below a submerged sluice gate, M.Sc. Thesis, Univ. of Minnesota, USA. Umbrell, E.R., Young, G.K., Stein, S.M. and Jones, J.S. (1998), Clear-water contraction scour under bridges in pressure flow, Journal of Hydraulic Engineering, 124 (2), Valentin, F. (1967), Considerations concerning scour in the case of flow under gates, Proc. 12 th Cong. IAHR, Vol. 3, Wu, S. and Rajaratnam, N. (1996), Submerged flow regimes of rectangular sharp-crested weirs, Journal of Hydraulic Engineering, 122(7), Wu, S. and Rajaratnam, N. (1998), Impinging jet and surface flow regimes at drop structures, Journal of Hydraulic Research, 36(1),